Enhanced Harmony Search Algorithm For Better Decision Tree Classification Of Huge Medical Data

Shibghatullah, Abdul Samad and Abal Abas, Zuraida and Mohamed Aminuddin, Mai Mariam and Asmai, Siti Azirah and Basiron, Halizah and Mustaffa, Izadora and Abdul Rahman, Ahmad Fadzli Nizam and Mansor, Nur Farraliza (2017) Enhanced Harmony Search Algorithm For Better Decision Tree Classification Of Huge Medical Data. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

Text Enhanced Harmony Search Algorithm For Better Decision Tree Classification Of Huge Medical Data.pdf - Submitted Version Restricted to Registered users only Download (285kB)

Abstract

As a meta-heuristic algorithm,Harmony Search (HS) algorithm is a population-based meta-heuristics approach that is superior in solving diversified large scale optimization problems.Several studies have pointed that Harmony Search (HS) is an efficient and flexible tool to resolve optimization problems in diverse areas of construction,engineering,robotics,telecommunication,health and energy.In this respect,the three main operators in HS,namely the Harmony Memory Consideration Rate (HMCR), Pitch Adjustment Rate (PAR) and Bandwidth (BW) play a vital role in balancing the local exploitation and the global exploration.These parameters influence the overall performance of HS algorithm,and therefore it is very crucial to fine turn them.However,when performing a local search, the harmony search algorithm can be easily trapped in the local optima.Therefore,there is a need to improve the fine tuning of the parameters.The contributions to this research are the Harmony Memory Consideration Rate (HMCR) parameter using step function while the fret spacing concept on guitars that is associated with mathematical formulae is also applied in the bandwidth (BW) parameter.The experimental results revealed that our proposed approach is superior to other state of the art harmony search known as parameter adaptive harmony search (PAHS).In conclusion,this proposed approach managed to generate a better results.

Actions (login required)

View Item